Towards an Optimization Pipeline for the Design of Train Control Systems with Hybrid Train Detection

Increasing the capacity of our railway infrastructure will become more and more essential in coping with the need for sustainable transportation. This can be achieved by intelligently implementing train control systems on specific railway networks. Methods that automate and optimize parts of this planning process are of great interest. For control systems based on hybrid train detection, such optimization tasks simultaneously involve routing and block layout generation. These tasks are already complex on their own; hence, a joint consideration often becomes infeasible. This work-in-progress paper proposes an idea to tackle the corresponding complexity. To this end, we present a pipeline that allows to sequentially handle corresponding optimization tasks in a less complex fashion while generating results that remain (close to) optimal. Results from an initial case study showcase that this approach is, indeed, promising. A prototypical implementation is included in the open-source Munich Train Control Toolkit available at https://github.com/cda-tum/mtct.